5. Potential travel improvements
5.2 Changing car fleet usage: scope and approach
I
NTRODUCTIONThis section discusses the methods used to find a number of trips that can be improved and to quantify the bene- fits of changing the way employee travel needs are satisfied. Five opportunities for travel need satisfaction are proposed. Each travel need satisfaction opportunity is reviewed in isolation (thus without considering the effects of the other opportunities). This was done to better compare the effects of all different opportunities for travel need satisfaction. First a number of boundary conditions are proposed. These boundary conditions represent (changes in) quantities of time and distance, for a number of transport modes, that are deemed acceptable in order to still comfortably travel for business purposes.
S
COPEThescopeofthisstudy(chapter1)specificallydoesnotfocusonthewillingnessofemployeestochangetheirtravel behaviour.However,theeffectivenessofchangingcarfleetusagedependsonthewillingnessofemployeestotake detoursofacertainlength,ortravelusingdifferenttransportmodes.Therefore,anumberofboundariesfordetour length/durationandmaximumwalkingandcyclingdistancesareset.Theseboundariesareprimarilysetbasedon reasonabledistancesanddurationsthatpeopletravelusingthespecifictransportmodes.Theboundariesarealso designedbasedontheassumptionthatanincentivesystemissetinplacebyCapgeminitoencourageemployeesto adjusttheirtravelbehaviour.
Car
Inordertomoreefficientlysatisfythetransportneedofallleasecarusingemployees,itisassumedthatindividual leasecaruserscanbepersuadedtotakeadetourfromtheirquickestroutetotheirdestination.Adetourof5kilome- tresmeasuredfromthetriporiginortripdestinationisdeemedthemaximumallowabledetourthatcanbeexpected fromaleasecaruser.Themaximumallowabledetourissetbasedonamaximumtraveltimedelayof36minutes,as describedbelow.
5. Potential travel improvements
FIGURE 26 EXAMPLE OF DETOUR KILOMETRES POSSIBLY MADE BY CAPGEMINI LEASE CAR USERS IN ORDER TO FACILITATE TOTAL EMPLOYEE TRAVEL NEED.
In the worst case scenario, the detour results in the lease car user is travelling 20 kilometres more than he would have when not taking a detour. Assuming this detour is taking place during rush hour in the Netherlands, then the worst possible average travel speed equals 34 km/hour15 (N213 near The Hague, INRIX, 2014). Thus the worst possible detour imaginable (20 km’s), at the worst possible time, at the worst possible location, would result in a trip delay of
⁄ . This worst case scenario is the limit of the average time delay an
employee can expect to incur when taking a detour. However, this being a worst case scenario, most detours will take a shorter amount of time then the above mentioned extreme.
Beside an increase in travel time, it is also assumed that employees are willing to adjust their time of arrival at their destination in order to more efficiently satisfy the transport needs of the entire lease car using population. It is assumed that employees are willing to arrive up to 15 minutes earlier or 15 minutes later at their destination in order to facilitate someone else’s travel need16
. Changing the preferred time of arrival does not necessarily result in a travel time delay because these 15 minutes can be used productively in one way or another.
Bicycle
According to a survey of the Central Statistical Office (CBS) Dutch commuters prefer using the bicycle when commuting distances up to 12.5 kilometres (single journey) (CBS, 2001). It is assumed that Capgemini employ- ees are willing to travel up to 12.5 kilometres single journey using a bicycle.
Every trip shorter than 12.5 kilometres is classified as performable by bicycle. However, when numerous busi- ness trips are performed during the day (all shorter than 12.5 kilometres), this could result in a situation where an employee has to travel hundreds of kilometres per day using the bicycle. In order to prevent this, a limit of 36 bicycle kilometres per day is set. The average bicycle velocity is 18 kilometres per hour (fietsersbond.nl, 2013), thus this limit results in a maximum bicycle time of 2 hours per day.
No differentiation is made between usage of a normal bicycle and an E-bike in this study. Further research taking into account the distances people are willing to travel by E-bike is recommended.
Walk
The average walking speed of pedestrians (ages 14 to 64) is 4.5 kilometre per hour (Knoblauch, Pietrucha, & Nitzburg, 1996). It is assumed that employees can be persuaded to walk distances up to 1.5 kilometres (single journey). This results in a maximum walking time of 20 minutes per walking instance.
15 Of course worse travel times are possible when extreme weather or accidents cause additional delays. The stated average speed is based on regular traffic conditions.
16
For future research, it is recommended to measure the actual willingness of Capgemini employees to wait for a carpool using, for example, a survey.
A
PPROACH:
CARPOOLINGWhy consider this change in satisfying transport need?
The first of five opportunities for changing the way travel needs are satisfied that are considered, is promoting employees to carpool when possible. Employees travelling from roughly the same location to a new location at roughly the same time can possibly share a car, instead of travelling to the new location with two (or more) cars. For each carpool, one lease car is driven less which is beneficial from a cost and environmental perspective.
How are the trips identified that could potentially be performed differently?
In order to find possible carpool opportunities, each registered trip in the dataset is compared to every other trip in the dataset. When two trips depart within a radius of 5 kilometres of each other, arrive at destination within a radius of 5 kilometres of each other, and arrive at the destinations within 15 minutes of one another (as defined in the scope discussed above), the two trips are classified as carpoolable. Additionally, the origin and destination of both trips should be located in the Netherlands. International trips are excluded because the exact locations (origin and destination) of those trips are less accurate then the trips located in the Netherlands (chapter 4). Be- cause very little trips were performed outside of the Netherlands, excluding international trips is expected to have little effect on the total number of potential carpool trips. Trips that do not leave their aggregated city (see appendix F for definition of aggregated cities) are also excluded from possible car sharing. This was done to prevent unrealistic care sharing17. In order to quantify the effects of car sharing, it is assumed that one car can transport a maximum of two persons18. The car that drove the smallest distance from origin to destination is then removed from the total travelled kilometres (figure 27). Only trips performed during work days are taken into account. This was done in order to more accurately represent the carpool opportunities when aggregated to aver- ages per (work) day19. Excluding weekend trips is expected to affect the total number of possible carpools very little due to the limited business trips performed during weekends (chapter 3). Performing the described carpool analysis will result in a rough estimate of the carpool opportunities for the Capgemini lease car users.
FIGURE 27 EXAMPLE OF CALCULATION FOR FINDING POSSIBLE CARPOOL OPPORTUNITIES AND THE RESULT- ING REDUCTION IN CAR KILOMETERS.
17
An example of unrealistic car sharing: An employee travels 4 kilometres by car from north to south Utrecht. Because of the small distance travelled, his destination is located in the 5 kilometre radius of his origin. Similar- ly, another employee is travelling 4 kilometres from south to north Utrecht. When trips in the same city are not excluded, these two employees would be counted as a possible car share opportunity, whereas they are actually travelling in opposite directions.
18
Thus when three trips are found that are travelling from roughly the same origin, to roughly the same destina- tion, at roughly the same time, the number of trips needed to transport these employees is only reduced by one. When four trips are found (ceteris paribus), then the number of needed trips are reduced by two.
19
When the carpool opportunities are aggregated for all days of the week including the weekends, significantly lower carpools per day are found because of the reduced business trips during weekends (chapter 3).
5. Potential travel improvements
The above described method for finding potential carpools might not give an accurate image of the actual car- pools that would occur after encouragement from Capgemini management (financially, and/or by other means). It is hypothesised that a certain level of predictability is required in order for employees to logistically arrange a carpool at the location and time of their preference. To account for this, the potential carpools are filtered. It is assumed a carpool is only logistically possible when the specific carpool from A to B occurs either on a weekly basis three times in a row, or three times within 6 working days, with a maximum of 1 day in between (excluding weekends; figure 28).
FIGURE 28 EXAMPLE OF POSSIBLE CARPOOL ACCEPTED (CAR 1 & 2) AND REJECTED (CAR 3 & 4) BECAUSE OF PREDICTABILITY AND UNPREDICTABILITY RESPECTIVELY.
Moreover, finding a carpool from A (home) to B (work) does not guarantee a carpool possibility from B (work) to A (home) at the end of the day. Because employees tend to prefer to get back home at the end of the day, a transport possibility back home is most likely a prerequisite for an employee to consider a carpool. Moreover, when an employee has to travel for business beside commuting, sharing a car might result in employees not being able to travel to their next meeting. To take this into account, all travel needs of an employee during the day are taken into account. Carpools are only deemed feasible when all employees involved can satisfy all their travel needs during the day under consideration using a carpool. Furthermore, at the end of the day, the employee should be within 2 kilometres of its starting point at the beginning of that day. This is necessary because the employee should be able to use his own car the next day.
FIGURE 29 EXAMPLE OF POSSIBLE CARPOOLS ACCEPTED (CAR 1 & 2) AND REJECTED (CAR 3 & 4) BECAUSE OF THE ABILITY TO GET BACK HOME AT THE END OF THE DAY (CAR 1 & 2) OR NOT (CAR 3 & 4).
Three different scenarios for car sharing are thus under consideration:
Scenario name Assumptions
Carpool rough
Assume an application exists that notifies Capgemini employees when another employee is leaving and from what location. This application can then be used to arrange a carpool ad hoc. When a car pool is found to travel from the home to the workplace but not vice versa, the em- ployee uses public transport, to get back home.
Carpool recurring trips
No applications exists that notifies Capgemini employees about possi- ble carpools. Therefore, carpool possibilities that are recurring and predictable are only feasible. When a carpool is arranged to travel from the home to the workplace but not vice versa, the employee uses public transport, to get back home.
Carpool getting home
An application exists that notifies Capgemini employees about a possi- ble carpool. Alternatively, Capgemini employees rely on informal collaboration to arrange their carpools ad hoc. When a carpool is arranged to travel from the home to the workplace but not vice versa, the carpool is disqualified. Employees have little trust that public transport can get them back home.
TABLE 8 CARPOOL SCENARIOS
How are the benefits of this transport change quantified?
The effects of the different carpool scenarios are calculated using indicators for the cost reduction and carbon dioxide emission per kilometre driven. The Director Facilities of Capgemini, Jack Knol stated that each kilome- tre driven by a Capgemini car results in a 0.109 euro expenses for Capgemini20. This key figure is used in com- bination with the estimated reduction in kilometres driven (figure 27) to quantify the financial benefit for Capgemini of car sharing. The average carbon dioxide emission of the Capgemini lease car fleet is tracked by Capgemini. The average Capgemini lease car exhausts little over 140 grams of CO2 per driven kilometre (Knol, 2015). It is assumed that every kilometre not driven by the lease car fleet therefore results in a carbon dioxide reduction of 140 grams.
Out of the roughly 3000 Capgemini lease cars, only 1729 are observed travelling during the 8 months for which trips data was available. Thus, only little more than half of the lease cars are analysed. Not all lease cars were found in the dataset because not all lease cars are outfitted with a trip tracking device yet. In order to correct the possible financial and environmental benefits for this, the expected benefits are extrapolated from 1729 to 3000 lease cars21. Moreover, because the dataset only contains all trips performed during 8 months, the expected bene-
20 The savings for Capgemini, due to a reduction in the driven kilometres, consist of an increased residual value of the lease cars and a reduction in fuel costs. Note that the total costs of the lease car consist of the costs paid by Capgemini (roughly 0.109 euro per kilometre) and the costs of the lease contract subtracted from the employee mobility budget.
21
Extrapolation done by multiplying the cost reduction and CO2 emission reduction by
.
5. Potential travel improvements
fits are also extrapolated to a full year22. This way, an estimation of the expected benefits over a whole year for the entire lease fleet is found. These extrapolations are also done for all other opportunities to change travel needs satisfaction discussed below.
A
PPROACH:
TRAVEL BY TRAINWhy consider this change in satisfying transport need?
More intensive use of the train for business travels is considered as a possible way to change the way travel needs are satisfied. It is hypothesised that using the train is cheaper and more environmentally friendly then using a lease car. Employees who live and work near a train station might be persuaded to travel by train instead of using their lease car.
How are the potential trips identified that are eligible for change in transport?
The geographical location of all train stations in the Netherlands as well as the distance (by rail) between each station combination is available publically23 (rijdendetreinen.nl, 2016). This data is used to identify lease car trips that could have potentially been performed by train. The distance of the origin and destination of each trip in the lease car dataset is compared to the location of all train stations in the Netherlands. When both the origin and destination are located within a radius of 1.5 kilometre (walking distance) of a train station, the trip is deemed performable by train. Similar to the carpool potential, only trips performed during weekdays are consid- ered. Because the accuracy of the trip origin and destination outside of the Netherlands are questionable, and only information about the Dutch train stations are available, only domestic business trips are considered when looking for trips potentially performable by train.
Similar to finding potential carpools, the potential train trips should also take into account that employees need to be able to reach every destination they want to travel to during a business day. The employees should also be able to return to their car at the end of the day. Therefore, the potential train trips are filtered similar to the poten- tial carpool trips (figure 29).
Scenario name Assumptions
Train rough Employees make use of the train when possible. When a train ride is
not possible, a car is available to them to satisfy their travel need. Train getting home
When it is possible to travel from the home to the workplace by train but not vice versa, or when one of the non-commuting trips is not performable by train, all potential train trips during that day are dis- qualified.
TABLE 9 TRAIN SCENARIOS
How are the benefits of this transport change quantified?
Quantifications of the benefits of using the train instead of the car consists of two parts: the beneficial effects of using the lease cars less, and the disadvantageous effects of using the train instead. It was already discussed that every reduction in driven kilometre results in a 0.109 euro cost saving and 140g less CO2 emission. The emission of carbon dioxide per driven kilometre per person in the train is dependent on the occupancy rate of that specific train. In the past, the Nederlandse Spoorwegen (NS) has stated several figures about the carbon dioxide emission per kilometre of their trains. However, due to some controversy of the calculation method, no official figures are given by the NS anymore (AMT, 2015; NS, 2016). When taking the average occupancy rate of the train into account, on average a train exhausts 50 grams of CO2, according to the NS24 (AMT, 2015). This key figure is used to estimate the exhausted CO2 when a trip is performed by train instead of a lease car. The costs of a train ticket in the Netherlands is based on the distance travelled by train. The costs of the tickets per kilometre trav-
22 Extrapolation done by multiplying the cost reduction and CO
2 emission reduction by
.
23 Because the source of this data is somewhat unclear, the data was validated using Google Maps and the NS travel planner. Validation confirmed the accuracy of the data.
elled go down for each subsequent kilometre25. Because the exact distances between stations are known, the costs of the ticket when taking the train instead of the car can be calculated.
A
PPROACH:
TRAVEL BY BICYCLEWhy consider this change in satisfying transport need?
Taking the bike instead of using the car is considered as the next opportunity for changing car fleet usage. The carbon dioxide emission of using a bicycle is negligible when lifecycle emission is ignored (CO2 emission as the result of producing the bike) (Auchapt, 2013). Moreover, cycling has a net benefit for your health (De Hartog et al., 2010). Finally, by taking the bike, the cost of transportation is reduced by a significant amount, as will be discussed below.
How are the trips identified that could potentially be performed differently?
Similar to the previous opportunities for changing car fleet usage, only trips performed during work days are considered as a potential bicycle trip. Limits for cycling distance and duration each day are used as described in the scope discussed previously. Only trips to and from a location in the Netherlands are considered in order to prevent forcing an employee to take his bike with him when driving to e.g. Italy for a business meeting.
Two scenarios for using the bicycle are considered: a rough scenario and a scenario that takes into account all trips that are performed during the day (similar to the train scenarios).
Scenario name Assumptions
Bicycle rough